Influence of the upper critical field anisotropy on the transport properties of

polycrystalline MgB 2

M. Eisterer/'Q C. Krutzler, 1 and H. W. Weber 1

1 Atominstitut der Osterreichischen Universitaten, 1020 Vienna, Austria
(Dated: February 2, 2008)

The intrinsic properties of MgB2 form the basis for all applications of this superconductor. We
wish to emphasize that the application range of polycrystalline MgB2 is limited by the upper critical
field H C 2 and its anisotropy. In wires or tapes, the MgB2 grains are randomly oriented or only
slightly textured and the anisotropy of the upper critical field leads to different transport properties
in different grains, if a magnetic field is applied and the current transport becomes percolative.
The irreversibility line is caused by the disappearance of a continuous superconducting current path
and not by depinning as in high temperature superconductors. Based on a percolation model, we
demonstrate how changes of the upper critical field and its anisotropy and how changes of flux
pinning will influence the critical currents of a wire or a tape. These predictions are compared to
results of neutron irradiation experiments, where these parameters were changed systematically.

PACS numbers: 74.70.Ad, 74.81.Bd, 74.25. Sv, 64.60. Ak

INTRODUCTION

DISTRIBUTION OF THE UPPER CRITICAL
FIELD

The strong field dependence of the critical current den-
sities in MgB2 represents one of the major problems for
power applications of this material, where a current den-
sity of at least 10 8 A/m 2 is needed. Such high current
densities can be obtained in conventional superconduc-
tors up to fields close to the upper critical field. In high
temperature superconductors thermally activated depin-
ning limits high current densities to relatively small mag-
netic fields compared to the huge upper critical fields of
these materials. In MgB2 thermal effects should play only
a minor role due to its comparatively low Ginzburg num-
ber. Nevertheless, the critical currents become too small
for power applications at fields well below H C 2, even at 4.2
K. Fortunately, the upper critical field of MgB 2 can be
rather easily enhanced by certain preparation conditions
QUI!, doping 0,H EiHH or irradiation HBHE1,
and exceeds 30 T (at K). Even for such "high-H C 2" ma-
terials, the application range is limited to around 10 T
in p olycrystalline samples. It was pointed out recently
[12j that this small application range was caused by the
upper critical field anisotropy of MgB 2 . In the pure ma-
terial, B c2 is about 14 T, if the field is applied parallel to
the boron planes, while it is smaller than 3 T, if the field
is applied perpendicular to the boron planes, leading to
an anisotropy of close to 5 [IE O EE EE E3, 13 In
polycrystalline MgB2, the grains are randomly oriented
and the first grains become normal conducting at B^,
which is a few times smaller than the upper critical field
of the whole sample (bJ,' 2 ), thus reducing the effective
cross section and the critical current of the conductor.
In this paper the influence of the anisotropy will be dis-
cussed quantitatively. We will demonstrate that B^ is
most important for the application range of MgB2.

The material is assumed to be perfectly homogeneous,
i.e. all grains are supposed to have identical intrinsic
properties. B C 2 of each grain depends only on its orien-
tation to the applied field and can be calculated within
anisotropic Ginzburg Landau theory |l9|:

B c2 {0)

B

^/ 7 2 cos 2 (6>) + sin 2 (6>)

(1)

7 denotes the anisotropy factor of the upper critical field,
i.e. B^ 2 /Bc2, and 9 is the angle between the applied field
and the c-axis. This relation was found to be a good ap-
proximation for MgB2 by torque measurements [Tj, Il5j ,
although deviations by a few percent were reported [20j.
For a given angular distribution of the grains, the distri-
bution of B C 2 can be easily calculated from Equ. In
untextured MgB2 , the grains are randomly oriented (each
crystallographic direction is equally distributed over the
elements of the steradian dtt — sin 9d9d<fi) and their an-
gles 9 with respect to any direction are distributed as
sin 9. Assuming b\ 2 to be 14 T and the anisotropy 7 to
be 4.5, as in the pure material, the distribution of Fig.
is obtained. Unfortunately, more grains have a compar-
atively low B C 2- The material can now be considered as
inhomogeneous, since the grains have different upper crit-
ical fields, depending on their orientation to the applied
field. This field induced inhomogeneity can be observed
by the broadening of the resistive transition in magnetic
fields (Fig. [2J. At zero field, where the upper critical
field plays no role, the transition of the filament of an
iron sheathed wire [2l| is sharp and its width (0.7 K) not
much larger than the transition width (0.3 K) of a sin-
gle crystal. At higher fields, the anisotropy significantly
broadens the transition of the wire (7.5 K) compared to

2

0.40

FIG. 1: Distribution of the upper critical field within the
grains

T(K)

FIG. 2: Resistive transition of a poly crystalline sample ("fil-
ament") and of a single crystal (H||a6). The applied current
density was in both specimens about 10 s A/m 2 .

the single crystal (1.8 K).

RESISTIVE TRANSITION

If a polycrystalline sample is cooled in a fixed applied
field Bq, the resistivity starts to deviate from its normal
conducting behavior, when the first grains become super-
conducting. B C 2 is defined by this onset of the transition
and, therefore, Bq corresponds to BjJ 2 at the actual tem-
perature T on . With decreasing temperature more and
more grains become superconducting and the resistivity
decreases. A grain becomes superconducting, if its up-
per critical field at the actual temperature becomes larger
than the applied field. For a continuous superconduct-
ing current path, a certain fraction of the grains has to

be superconducting. This fraction is called percolation
threshold or critical probability p c and depends on the
number of connections between the grains (coordination
number). Percolation theory predicts a lower limit for
p c of about 0.162 for an irregular lattice ("Finney
pack" ) with a coordination number of 14.3. If the coordi-
nation number is only 6, p c is calculated to be about 0.31
for a three dimensional lattice. These values correspond
to the site percolation problem, which is the appropriate
model for MgB2, since the nodes (grains in the real sys-
tem) are switched on or off in site percolation problems,
while the bonds between the nodes (grain boundaries) are
switched in bond percolation problems. The fraction of
superconducting grains p at a field Bq , can be calculated
by integration of the distribution function:

n_

2

p= J sin 6d6 = cos 9 min (2)

9 min

In the transition region Q m in is defined by B C 2.(B m in) =
Bq. Inserting relation [3 with p = p c into Equ. pleads to
the field of zero resistivity

B p=0 (T) = n jifi i (3)

The field, where the resistivity disappears, is commonly
denoted as the irreversibility field. Since the finite re-
sistivity in MgB 2 above B p=0 is caused by a completely
different mechanism than in high temperature supercon-
ductors, this field is denoted as "zero resistivity field" in
the following. According to Equ. [3] the zero resistivity
field is proportional to the upper critical field, as ob-
served in experiments (Fig. |3J. The symbols represent
experimental data obtained on the filament of an iron
sheathed wire. The onset and the offset of the transition
were defined by 90 % and by 10 % of the resistivity at
40 K, respectively. The line graph was calculated with
7 = 4.5 and p c = 0.25.

At fixed applied field the condition for the offset of
the transition becomes B p= o(T°^) = B^ and - assuming

only B^ 2 to be temperature dependent - the transition
width AT := T on - T°H is obtained as [13

.^yfEpzl^ (4)

\ dT I

The transition width is predicted to decrease with de-
creasing anisotropy 7. This can be demonstrated by
neutron irradiation experiments. Neutron irradiation in-
creases the upper critical field of MgB2 and decreases its
anisotropy |23l . The resistive transition of a sintered
MgB2 sample is plotted in Fig. ^before and after neutron
irradiation. The shift of B C 2 and the predicted decrease
of the transition width can be observed.

3

2 4 6 8 10 12 14

FIG. 3: Zero resisitivity field as a function of the upper critical
field. Symbols refer to experimental data, the line represents
the expected behavior for 7=4.5 and p c =0.25.

6 8 10 12 14 16 18 20 22

T(K)

FIG. 4: Resistive transition at 13 T before and after neutron
irradiation.

2.5

5 10 15 20 25

T(K)

FIG. 5: Resistive transition of a wire before and after remov-
ing the conducting sheath

sheath can decrease the experimental transition width
significantly, as shown in Fig. [5] The iron sheath of a
wire with a total diameter of 1.1 mm was removed by
etching, the remaining filament had a diameter of about
0.65 mm. The resistivity increased by nearly two orders
of magnitude from about 2.28 /ificm to about 160 /lilcm.
The onset of the transition (defined by the 90 % crite-
rion) was shifted from 12.7 K to 20 K and the transition
width increased from 3.8 K to 7.5 K. These changes are
mainly due to the removal of the parallel resistivity of
the iron sheath, since the original curve can be recal-
culated by assuming a parallel resistivity of 2.32 //ficm
(dotted curve in Fig. |SJ). The small difference can be
explained by the resistivity between the iron sheath and
the filament and/or by small property changes due to the
strain exerted by the iron sheath (2^ . For the analysis of
the anisotropy, the real B C 2 and B p= o axe needed, which
can only be determined after removing the conducting
sheath.

In our considerations we neglect the influence of sam-
ple inhomogeneities and of thermal activation of the flux
lines. Both effects would additionally increase the tran-
sition width. Inhomogeneities shift the transitions of
equally oriented grains and thermally activated depin-
ning broadens the transition of each individual grain,
but is assumed to be zero (cf. the transition of the sin-
gle crystal in Fig. [3J . However, in reasonably homoge-
neous samples, the anisotropy is the dominant parameter,
since it predicts the correct magnitude of the transition
width (except for low magnetic fields). The transition
width calculated from Equ. 0] represents a lower limit
and is close to typical experimental data under realis-
tic assumptions for p c and for 7, not leaving much room
for additional effects. On the other hand, a conducting

CRITICAL CURRENTS

From the distribution of the upper critical fields (cf.
Fig. ^| the distribution of the critical current densities
within the grains can be obtained, if a certain pinning
model is assumed. For grain boundary pinning, J c is
related to B C 2 via [2^|

m - jJ l - BIB ^ e))2 ( 5 )

^B c2 {6)B

for B < B C 2 and J c = otherwise. The coefficient J c o
represents the pinning strength. J c of a polycrystalline
sample can be calculated from the distribution of the
critical current densities within the framework of
percolation theory. This is demonstrated in Fig. for

4

A

□

O

io 10 j

10 9 [

CVI

10 8 [

<

10 7 [

—3

10 6 [

10 5 [

10 4 :

J c0 (10 9 A/m 2 )
bulk 3.2
wire Fe 20
wire Cu 3.3

B c2 TO

Y

Pc

21.4

4.4

0.21

25.5

4.4

0.23

15.4

4.4

0.26

B(T)

FIG. 6: Critical current densities of a bulk sample at 5 K and
of two different wires at 4.2 K.

different samples, i.e. a sintered bulk material, a copper
sheathed in-situ wire |27l | and an iron sheathed in-situ
wire [2]]] ■ The experimental data (symbols in Fig.
were obtained by direct transport measurements (using
a 1 /iV/cm criterion) in the case of the wires, and by ac
measurements |28| on the bulk sample. For the calcula-
tion of J c (lines in Fig. EJ) four parameters are needed.
The upper critical field B C 2 of the sample (i.e. b\ 2 of
the grains), the anisotropy 7, the percolation threshold
p c and the pinning strength J c q. While B C 2 was mea-
sured directly at high temperatures and extrapolated to
low temperatures, the other parameters have to be fitted
to the experimental data. The critical current densities
can be well described by the same anisotropy for all three
samples and similar values for the percolation threshold.
The pinning strength of the Cu-shcatcd wire and the bulk
sample is nearly identical, the smaller field dependence
of J c in the bulk sample is caused by its higher B C 2 . The
iron sheathed wire has the largest upper critical field and
shows the strongest pinning, which leads to the highest
current densities at all fields.

The primary infiucence of neutron irradiation in MgB2
lies in the increase of B C 2 and the decrease of anisotropy.
Nearly no changes of J c at low fields, but a strong
enhancement of J c at high magnetic fields are ob-
served (Fig. . The upper critical field increases from
21.4 T and 15.4 T to 30 T and 18.6 T, respectively,
the anisotropy 7 (obtained from fitting) decreased from
about 4.4 in both samples to 2.8 and 2.6 in the bulk
and in the copper sheathed wire, respectively. This de-
crease of 7 is consistent with direct measurements of the

-| — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — r

before/after irradiation
A / A bulk
° / • wire Cu -

2 4 6 8 10 12 14 16 18
B(T)

FIG. 7: Critical current densities before and after neutron
irradiation.

upper critical field anisotropy in single crystals |23J, |2J
The influence on pinning is rather small. The increase
of B c2 reduces the pinning strength [2g (J c cx 1/y/B^),
but J c o of the sintered bulk sample increases by 25 %
after irradiation, which can be attributed to the intro-
duction of pinning centers, as observed in single crys-
tals [29^. These competing effects lead to an increase of
the pinning strength by only about 5 % at low magnetic
fields. The pinning strength even decreases by about 10
% in the wire. Therefore, the enhancement of J c after
neutron irradiation is caused mainly by changes in the
reversible parameters of MgB 2 . Since these changes of
the reversible parameters were observed independently
on single crystals, the percolation model 01 correctly
predicts the changes of the critical currents in polycrys-
talline samples.

CONSEQUENCES / STRATEGY

Based on the quantitative agreement between
anisotropy-induced percolation theory and experimental
data on polycrystalline MgB2, the fundamental parame-
ters of this theory can be exploited for an assessment of
their individual impact on the current carrying capability
of MgB2. We will, therefore, systematically investigate
the influence of variations in its four parameters. J C (B)
calculated from the parameters of the iron sheathed wire
will be used as reference (dotted line in Fig. |HJ and only
one parameter will be changed by 30 % in each scenario.
The critical current density of the wire drops to 2 x 10 8
A/m 2 (horizontal lines in Fig. |SJ), which is a reasonable
limit for power applications and defines the application
range in the subsequent investigation, very closely to the
upper critical field B^ = 5.8 T (vertical lines in Fig. |HJ.

The first strategy is an improvement of pinning. This
can be achieved by the addition of pinning centers or,

in the case of grain boundary pinning, by a very fine
grain structure. If the corresponding parameter, J c q, is
increased, the critical current densities are enhanced at
all fields by the same factor (Fig. [SJi) . Although this is
favorable at low magnetic fields, it has little effect on
the application range. The simple shift (on a logarith-
mic scale) of J C (B) to higher values only occurs, if the
pinning mechanism remains unchanged, i.e. the critical
current densities of the individual grains can be described
by Equ.0 However, since J c becomes zero at B c2 for any
pinning mechanism, the strong field dependence and the
importance of B^r 2 cannot be changed qualitatively by
different pinning centres.

As a second strategy, an increase of the upper crit-
ical field is considered. It is well documented that
Bp2 can be enhanced significantly by various techniques
Ef H SSHESHSIliillll Although the mechanisms
involved in these changes are not yet fully understood,
impurity scattering in both bands seems to play a major
role. However, there must be a mechanism, which en-
hances B\ 2 without significantly changing 7 (cf. Fig. HJJl .
A shift of the upper critical field from 25.5 T to 33.15 T
results in a strong increase of the zero resistivity field
and in higher critical currents at high magnetic fields
(Fig. |Ht>) ■ Since the anisotropy is constant, B^r 2 increases
to 7.5 T. J c falls below 2 x 10 s A/m 2 exactly at B^ 2 .

A smaller percolation threshold p c is the third possi-
bilty to improve the current carrying capability of poly-
crystalline MgB2. This can be achieved by a homoge-
neous sample with a high density. Pores and normal con-
ducting grains (e.g. MgO) increase p c . A small percola-
tion threshold enhances the zero resistivity field, but has
no influence on J c within the application range (Fig.|SJ;).
In principle the densitiy of the sample also affects J c q
due to the reduction of the superconducting cross sec-
tion, but this effect is small in samples of good quality.
For the present parameters, an increase of J c q by about
15 % is expected. Since a p c of 0.16 corresponds to the
theoretical limit |2^], a significant improvement of the
transport properties cannot be expected from the opti-
mization of p c .

The last strategy is a reduction of the anisotropy 7.
A reduction of 7 was reported after neutron irradiation
[23I Eif or can be achieved by carbon doping [3(J , but
it seems to be always accompanied by an increase of the
upper critical field B^ 2 (which further enhances B^r 2 and
enlarges the application range). Although experimen-
tally difficult, the influence of such a "pure" change of
anisotropy on J c is demonstrated in Fig. [SJi. 7 is re-
duced from 4.4 to 3.1 or, equivalently, Bjr 2 is increased
by 2.4 T (from 5.8 T to 8.2 T). This also enlarges the
application range by about 2 T and emphasizes again
the importance of B^r 2 for power applications of MgB2 in
magnetic fields.

1 I 1 I 1 I 1 I _

J =2x1 10 A/m 2 i

CO ;

J =2.6x1 10 A/mn

CO :

P c =0.23 =
P°=0.16 1

2 4 6 8 10 12 14 16

B(T)

FIG. 8: Influence of various parameters on J c . Only one
parameter of the original data set (dotted curve) is changed
in each panel.

6

CONCLUSIONS

The upper critical field and its anisotropy are the most
important and determining for the field of zero resistivity
in polycrystalline MgB2 , in contrast to high temperature
superconductors, where the irreversibility field is related
to the pinning properties. The maximum magnetic field
for power applications of MgB2 is always close to B^r 2 . ft
is, therefore, essential to enhance Bjr 2 for applications of
MgB2 at high magnetic fields.

We wish to thank Sonja Schlacher and Wilfried
Goldacker from the Forschungszentrum Karlsruhe for
providing us with the iron sheathed wire.

* Electronic address: eisterer@ati.ac.at
[1] Y. Takano, H. Takeya, H. Fujii, H. Kumakura, T.

Hatano, K. Togano, H. Kito, H. Ihara, Appl. Phys. Lett.

78, 2914 (2001).
[2] R. Fliikiger, P. Lezza, C. Beneduce, N. Musolino, H. L.

Suo, Supercond. Sci.Technol. 16, 264 (2003).
[3] A. Gurevich, S. Patnaik, V. Braccini, K.H. Kim, C.

Mielke, X. Song, L.D Cooley, S.D. Bu, D.M. Kim, J.H.

Choi, L.J. Belenky, J. Giencke, M.K. Lee, W. Tian,

X.Q. Pan, A. Siri, E.E. Hellstrom, C.B. Eom, D.C. Lar-

balestier, Supercond. Sci.Technol. 17, 278 (2004).
[4] R.A. Ribeiro, S.L. Bud'ko, C. Petrovic, PC. Canfield,

Physica C 384, 227 (2003).
[5] S.X Dou, V. Braccini, S. Soltanian, R. Klie, Y. Zhu, S.

Li, XL. Wang, D. Larbalestier, J. Appl. Phys. 96, 7549

(2004).

[6] R.H.T. Wilke, S.L. Bud'ko, P.C. Canfield, D.K.

Finnemore, R.J. Suplinskas, ST. Hannahs, Phys. Rev.

Lett. 92, 217003 (2004).
[7] S.M. Kazakov, R. Puzniak, K. Rogacki, A.V. Mironov,

N.D. Zhigadlo, J. Jun, Ch. Soltmann, B. Batlogg, J.

Karpinski, Phys. Rev. B 71, 024533 (2005).
[8] M.D. Sumption, M. Bhatia, M. Rindfleisch, M. Tomsic,

S. Soltanian, S. X. Dou, E. W. Collings, Appl. Phys. Lett.

86, 092507 (2005).
[9] M. Eisterer, M. Zehetmayer, S. Tonies, HW. Weber, M.

Kambara, N.H. Babu, D.A. Cardwell, L.R. Greenwood,

Supercond. Sci.Technol. 15, L9 (2002)
[10] Y. Wang, F. Bouquet, I. Sheikin, P. Toulemonde, B. Re-

vaz, M. Eisterer, H.W. Weber, J. Hinderer, A. Junod, J.

Phys.: Condens. Matter 15, 883 (2003).
[11] M. Putti, V. Braccini, C. Ferdeghini, F. Gatti, G. Grasso,

P. Manfrinetti, D. Marre, A. Palenzona, I. Pallecchi,

C. Tarantini, I. Sheikin, H. U. Aebersold, E. Lehmann,

Appl. Phys. Lett. 86, 112503 (2005).
[12] M. Eisterer, M. Zehetmayer, H.W. Weber Phys. Rev.

Lett. 90, 247002 (2003).
[13] Y. Eltsev, S. Lee, K. Nakao, N. Chikumoto, S. Tajima,

N. Koshizuka, M. Murakami, Phys. Rev. B 65, 140501

(2002).

[14] M. Angst, R. Puzniak, A. Wisniewski, J. Jun, S. M.

Kazakov, J. Karpinski, J. Roos, H. Keller, Phys. Rev.

Lett. 88, 167004 (2002).
[15] M. Zehetmayer, M. Eisterer, J. Jun, S. M. Kazakov, J.

Karpinski, A. Wisniewski, H.W. Weber, Phys. Rev. B

66, 052505 (2002).
[16] A. V. Sologubenko, J. Jun, S. M. Kazakov, J. Karpinski,

H.R. Ott, Phys. Rev. B 65, 180505 (2002).
[17] L. Lyard, P. Samuely, P. Szabo, T. Klein, C. Marcenat, L.

Paulius, K.H.P. Kim, C.U. Jung, H.-S. Lee, B. Kang, S.

Choi, S.-I. Lee, J. Marcus, S. Blanchard, A.G. M. Jansen,

U. Welp, G. Karapetrov, W.K. Kwok, Phys. Rev. B 66,

180502 (2002).

[18] U. Welp, A. Rydh, G. Karapetrov, W.K. Kwok, GW.

Crabtree, C. Marcenat, L. Paulius, T. Klein, J. Marcus,

K.H.P. Kim, C.U. Jung, H.-S. Lee, B. Kang, S.-I. Lee,

Phys. Rev. B 67, 012505 (2003).
[19] D. R. Tilley, Proc. Phys. Soc. (London) 86, 289 (1965).
[20] A. Rydh, U. Welp, A.E. Koshelev, W.K. Kwok, G.W.

Crabtree, R. Brusetti, L. Lyard, T. Klein C. Marcenat,

B. Kang, K.H. Kim, K.H.P. Kim, H.-S. Lee, S.-I. Lee,

Phys. Rev. B 70, 132503 (2004).
[21] W. Goldacker, S.L Schlachter, B. Obst, B. Liu, J. Reiner,

S. Zimmer, Supercond. Sci.Technol. 17, S363 (2004).
[22] S. C. van der Marck, Phys. Rev. E 55, 1514 (1997).
[23] M. Zehetmayer, M. Eisterer, J. Jun, S. M. Kazakov, J.

Karpinski, H.W. Weber, Phys. Rev. B 70, 214516 (2004).
[24] M. Eisterer, phys. stat. sol. (c) 2, 5, 1606 (2005).
[25] P. Kovac, M. Dhalle, T. Melisek, H.J.N, van Eck, W.A.J.

Wessel, B. ten Haken, I. Husek, Supercond. Sci.Technol.

16, 600 (2003).
[26] D. Dew-Hughes, Phil. Mag. 30, 293 (1974).
[27] B.A. Glowacki, M. Majoros, M. Vickers, J.E. Evetts,

Y. Shi, I. McDougall, Supercond. Sci. Technol. 14 193

(2001).

[28] M. Eisterer, H.W. Weber, J. Appl. Phys. 88, 4749 (2000).
[29] M. Zehetmayer, M. Eisterer, J. Jun, S. M. Kazakov, J.

Karpinski, B. Birajdar, O. Eibl, and H. W. Weber, Phys.

Rev. B 69, 054510 (2004).
[30] T. Masui, S. Lee, S. Tajima, Phys. Rev. B 70, 024504

(2004).